Journal of the Meteorological Society of Japan, Vol. 76 ...

Journal of the Meteorological Society of Japan, Vol. 76, No. 6, pp. 1045-1063, 1998 1045

Seasonal Change of Asian Summer Monsoon Circulation and Its Heat Source

By Tomoaki Ose

Meteorological Research Institute, Tsukuba, Japan

(Manuscript received 8 May 1998, in revised form 8 October 1998)

Abstract

The Asian summer monsoon circulation, especially its climatological seasonal change, was studied as the model response to the prescribed zonal mean field and the prescribed diabatic heat source from the observation. The obtained results are summarized as follows.

(1) During the Asian summer monsoon season, the prescribed deep heat sources in the southern part of Asia form the Tibetan High, the monsoon trough, the low-level circulation over South Asia, and furthermore, the downward motion in the western part of the Eurasian Continent. The heat sources near the surface over central Asia also induce downward motions aloft.

(2) In early summer (June), the deep heat sources in the southern part of Asia tend to form southwesterly low-level flows and upward motion southeast of Japan. Those are considered to be the background for the Baiu formation in East Asia as well as heat lows produced in the southern part of Asia. The

mid-latitude heat sources associated with the Baiu precipitation produce a low-level jet south of that.

(3) Climatological seasonal change from early summer (June) to mid-summer (July) is characterized by an air temperature increase in the whole Northern Hemisphere and a northward shift of a weakened westerly jet. When in the model a zonal mean field in June is replaced by that in July, the major characteristics of the seasonal change are obtained qualitatively; low-level jets and upward motion areas in South Asia and East Asia shift from the ocean side of the coasts toward the land side. This change of vertical motion is consistent with the seasonal change of deep heat sources from June to July.

(4) The climatological seasonal change from mid-summer (July) to late summer (August) is characterized by enhanced convective activity in the extended area of the subtropical western Pacific. When deep heat sources in July are replaced by those in August over the western Pacific only, the major characteristics of the seasonal change over the Pacific and the Indian Ocean are obtained. The expansion of the Tibetan High at the upper-level and the Pacific High at low-level over Japan is also simulated by the seasonal change of the western Pacific heat sources only.

(5) The model simulation with the combination of the diabatic heat source for August and the zonal mean field for June is compared with the climatological August simulation. It is indicated that the zonal mean field delayed from its seasonal migration could be related to weak monsoon circulation and the associated precipitation anomalies in the mid-latitudes and the subtropics.

1. Introduction

Hoskins and Rodwell (1995) show that many fea-tures of the Asian summer monsoon circulation av-eraged during June-August can be reproduced by a time-dependent primitive equation model with spec-ified tonal flow, mountains, and diabatic heating. Using the model, Rodwell and Hoskins (1996) dis-cussed the mechanism for the northern summer dry climate in the eastern Sahara, the Mediterranean, and the Kyzylkum desert. They show that the in-

teraction between the subtropical heat sources and the mid-latitude westerlies intensifies vertical mo-tion in the mid-latitudes. The extended downward motion in the mid-latitudes explains the existence of a dry region in the western Eurasian Continent. The purpose of this paper is to understand the

climatological seasonal change of the Asian summer monsoon circulation in a view of the model response to the prescribed heat sources (their wave compo-nents) and zonal mean fields, following Hoskins and Rodwell (1995). Apparently, the seasonal change of the tonal mean field is not independent from that of the heat sources, but other mechanisms are in-volved to maintain the zonal mean field. We can see

Corresponding author: Tomoaki Ose, Meteorological Re-search Institute, 1-1 Nagamine, Tsukuba, Ibaraki 305-0052, Japan. E-mail: [email protected]

1998, Meteorological Society of Japan

1046 Journal of the Meteorological Society of Japan Vol. 76, No. 6

the roles of heat sources and zonal mean fields in the seasonal change of the Asian summer monsoon cir-culation by exchanging tonal mean fields from one month to another and limiting the domains of the specified heat sources.

A wet climate is observed in East Asia and around Japan in the northern summer season. One of the most characteristic phenomena is the Baiu and the Meiyu (e.g., Ninomiya and Akiyama, 1992; Ding, 1992). The Baiu precipitation zone appears south of Japan in late May, then it migrates northward (Ninomiya and Murakami, 1987; Tanaka, 1992). The precipitation zone disappears around northern Japan in mid-July. Kodama (1992) pointed out that the Baiu, the South Pacific Convergence Zone (SPCZ), and the South Atlantic Convergence Zone (SACZ) have common features as subtropical pre-cipitation zones.

Furthermore, Kodama (1993) clarifies the condi-tions for the existence of the subtropical conver-gence zone. One of the conditions is that low-level poleward flows prevail in the western peripheries of the subtropical highs. Heat lows for producing the low-level poleward flow are formed by land surface heating and/or monsoon convection (Kato, 1989; Kodama, 1993). Another condition for the existence of the subtropical convergence zone is that the sub-tropical jets flow within almost 30-35N.

A question we need to ask is how the wet climate in East Asia is simulated by the model and how it is affected by the seasonal change of the zonal mean field from June to July. Ueda et al. (1995) showed that the abrupt sea-

sonal change occurs around late July in large-scale convective activity over the western Pacific. This maturing process in the western Pacific summer monsoon is studied as a coupled ocean/atmosphere system by Ueda and Yasunari (1996). Murakami and Matsumoto (1994) analyzed the western North Pacific summer monsoon in mid-August. They show that the mechanism for the western North Pacific summer monsoon is quite different from that for the Southeast Asian summer monsoon. According to the paper, the western North Pacific summer mon-soon is driven by the combined effects of zonal asym-metry in sea surface temperature (SST) at 10N to 20N, and continent-ocean heat contrasts between about 20N and 30N. We are interested in whether the seasonal change of the prescribed diabatic heating reproduces the seasonal change of the Asian summer monsoon cir-culation from mid-summer (July) to late summer (August). The method and the model are described in Sec-

tion 2. In Section 3, the June simulation is shown first, then the effects of regional heat sources are studied separately. The seasonal change from June to July is studied by comparing the June and July

simulations and presented in Section 4. Studies are concentrated to the effect of the tonal mean field change. In Section 5, the seasonal change from July to August is studied. We focus on the enhanced subtropical deep heating in the western Pacific in August. Conclusions are given in Section 6. 2. Model

The method to simulate the stationary summer monsoon circulation in this study is based on Hoskins and Rodwell (1995). The major part of the model is the dynamic part of the MRI-GCM-I (Tokioka et al., 1984), where horizontal and vertical

differences are based on Arakawa and Lamb (1977) and Tokioka (1978), respectively.

The model has 2.5x2.5 grids in longitudinal and latitudinal directions. It has seven layers vertically at 0.0555..., 0.222..., 0.444..., 0.638..., 0.777..., 0.888... and 0.972... in the modified or-coordinate;

or=(P-PT)/(PS-PT), (1)

where P and Ps are for the model-level pressure and surface pressure, respectively. PT is 100hPa. If P3 is 1000hPa, the model levels correspond to 150, 300, 500, 675, 800, 900, and 950hPa. The modified u- coordinate level is referred to as just sigma-level in this paper.

Prognostic variables are temperature (T)/poten- tial temperature (9), horizontal velocity (V) and surface pressure (PS). Simple terms are added to the dynamic frames as follows;

dT/dt=Q1-(T-Tm)/rT+(P/Po)R/cp}

[KhzO+aKvOB/az2], (2) dV/dt=-V/TV+KhdV+aKuaV/az2, (3)

where TT is a Newtonian cooling time of 25-day for 1-5 layers, 10- and 5-day for the bottom two layers,

Raleigh friction is applied to the second lowest layer and the lowest layer with a damping time of 5- and 1-day (5/4- and 1/4-day) over ocean (land), respec-tively. Vertical diffusion is applied to the top two layers with Kv of 16m2/s. Horizontal diffusion is included with a small coefficient Kh of 5.103m2/s for all layers.

Integration starts from zonal mean of the monthly mean T, V and P3, which are obtained from the ini-tialized ECMWF reanalysis data for 1985-87. Ac-cording to Hoskins and Rodwell (1995), zonal means of T, V and PS on the sigma-levels are kept fixed to those initial conditions throughout the model in-tegration, except the hydrostatic adjustments to T and PS. No mountain nor land is specified south of 60S. The prescribed heating rate Q1 is the monthly mean apparent heat source (Yanai et al., 1973). This is obtained by the atmospheric heat budget calcu-lation (Hoskins et al., 1989) from every 12-hour se-

December 1998 T. Use 1047

quence of the initialized ECMWF reanalysis data for 1985-87;

Q1=E(P/PO)(R/cp)(R/cP)

[zO/zXt+V.oe+Wae/aPJ/Cp}/E, (4)

where the usual notation is used. The symbol of represents the summation of each time step. Note that only the deviation of Q1 from its zonal aver-age is applied to the model. Although the period of 1985-87 is not very long for studying the clima-tological seasonal change, it will be shown that the climatological seasonal change of the precipitation for 1979-94 (Xie and Arkin, 1996) is well captured by that of the Q1 averaged during 1985-87, at least qualitatively. The mountains are lifted linearly with time dur-

ing the first five days of integrations while the hy-drostatic adjustment is made. The prescribed heat source is turned on day 5. The integration attains the equilibrium state almost by day 10. The simula-tion results shown are averaged during day 11-day 15.

A series of the model simulation presented in this paper are summarized in Table 1, where the symbolic names for the experiments, the prescribed zonal mean fields and the prescribed diabatic heat-ing are noted. The southern part of Asia, including South Asia,

the Bay of Bengal, Indochina, the South China Sea and the tropical western Pacific is referred to as south-east Asia hereafter.

3. Early Summer (June)

3.1 Heat sources and circulation Figure 1a shows the mass-weighted mean Q1 be-

tween 85hPa and surface for June. Large and pos-itive Q1 are found over south-east Asia. The Asian summer monsoon is characterized by convective heat sources at relatively high latitudes in the subtropics. A particularly large Q1 locates in the Bay of Bengal along 90E and near the Philippines. Precipitation due to the Baiu front (or the Meiu front) seems to expand northeastward from south-east Asia. Figure 1b shows the Q1 at the 0.89 sigma-level

in June. The Q1 between the 0.89 and 0.97 sigma-levels is hereafter referred to as near-surface heat-ing. Much of the near-surface heating over land is thought to be sensible heating. The near-surface heating is quite different from the vertically aver-aged heating distribution (Fig. 1a). Near-surface heating over 1K/day in the Northern Hemisphere is almost restrained to the area over the continents. Especially large near-surface heating is found over the Tibetan Plateau and the western and regions. The climatological precipitation for June during 1979-94 is shown in Fig. 1c. The features of the con-densational heating found in Fig. 1a are confirmed in the precipitation pattern. Figure ld shows the zonal mean zonal wind and

vertical motion prescribed for the June simulation with contours and arrows, respectively. The sub-tropical jet locates at 40N. Large vertical motions are found in the tropics and the southern subtrop-ics, but not in most of the northern subtropics and

Table 1. Symbolic names and contents of the experiments. The used zonal mean field and diabatic heating rate are

shown. See details in the text.


mid-latitudes. Figures 2a and 2b show the observed streamfunc-

tion at upper-level (200hPa) and low-level (0.78 sigma-level) for June. The vertical velocity at 500hPa in June is shown in Fig. 2c. Hereafter, 200hPa and 0.78 sigma-level are referred to as upper-level and low-level, respectively. The char-acteristics of the northern summer circulation are confirmed; the Tibetan High, the upper-level Pa-cific trough, the Indian monsoon trough, and the low-level westerly jet in South Asia. Upward mo-tion in south-east Asia, the Baiu region, west of the Caspian Sea, around the mountains on the western edge of the Tibetan Plateau, and downward motions in the western arid regions of the Eurasian Continent are observed. Figures 2d-2f are the same as Figs. 2a-2c ex-

cept for the simulated ones. The simulations are referred to as EXP-JUN. The observed large-scale features are well captured by the simulation. Down-ward motions are also simulated in the appropriate area, but those tend to be too large when compared with the observation. Too large a magnitude for the downward motion is also simulated by Rodwell and Hoskins (1996). They attributed it to poorly re-solved orographic features, such as the Zagros moun-

tains. Some small-scale features in the simulation are

magnified or not observed. This fact may be partly attributed to the lack of the horizontal hyper-diffusion term (Hoskins and Rodwell, 1995) in the model. A low-level cyclonic circulation over eastern Siberia is not found in the observation of Fig. 2b, but it may not be necessarily wrong in the simu-lation. A trough over this region, named the Baiu trough, is reported as an important factor for the Baiu front (Ninomiya and Muraki, 1986).

3.2 Response to subtropical deep heat sources Deep convection in south-east Asia is one of the characteristics in the Asian summer monsoon. We study the circulation created solely by the heat sources in south-east. Figures 3a-3c shows the sim-ulated streamfunction at upper-level and low-level, and 500hPa vertical velocity, respectively, for EXP-SOUTH where the June heat sources are specified only over the dark shaded region in Fig. 4 (referred to as the SOUTH region, hereafter). This region is subjectively chosen so as to include the major sub-tropical heat source but not the Baiu precipitation. The obtained circulations are quite similar to those of EXP-JUN within the SOUTH region. The heat sources in the SOUTH region reproduce the major

Fig. 1. (a) Vertically averaged diabatic heating rate (K/day) between 85hPa and surface for June. Dark and light shadings are for more than 0.5K/day and less than -0.5K/day, respectively. (b) Near-surface (see the definition in the text) heating rate (K/day) for June. Contours are every 2.0K/day. Dark and light shadings are for more than 1.0K/day and less than -1.0K/day, respectively. (c) Climatological precipitation rate (mm/day) from Xie and Arkin (1996). Contours are every 5mm/day. Shading is for more than 5mm/day. (d) The prescribed zonal mean zonal wind (m/s) and vertical velocity (0.01Pa/s) for June represented with contours and arrows, respectively.

Arrows are omitted for less than 0.01Pa/s in magnitude. The ECMWF reanalyses data is used for (a), (b) and (d).

(a) Q1 (*K/day) JUN (b) Q1 (SIG=0.89) (*K/day) JUN

(c) Precipitation (mm/day) JUN _(d) U(m/s) and W(s0.01Pa/s) ZMEAN JUN

December 1998 T. Ose 1049

characteristics of the Asian summer monsoon sim-ulation - the Tibetan High, the Indian monsoon trough, and the low-level westerly jet in South Asia. Responses outside the SOUTH region are inter-

esting. The outside responses in EXP-SOUTH have some similar features to those in EXP-JUN. Parts of the common features between EXP-JUN and EXP-SOUTH come from the interaction between the prescribed tonal mean fields and the topogra-phy. Figures 3d-3f show the topography effect on the upper-level, low-level circulation and 500hPa vertical motion in the EXP-TOPOG case where no diabatic heating is specified on the globe. The mountains produce downward motions over the east-ern Mediterranean Sea and the Tibetan Plateau, as shown by Rodwell and Hoskins (1996). Upward mo-tions are produced over the Caspian Sea, east of the Tibetan Plateau and east of the Sea of Okhotsk. Many of those topography effects are also confirmed in EXP-JUN and EXP-SOUTH. We can see two remote effects of the heat sources

in south-east Asia on the northern circulation out-side the SOUTH region by comparing EXP-SOUTH and EXP-TOPOG. One is the enhanced downward motions over the western arid regions, including the eastern Mediterranean Sea. This effect is one of the major conclusions of Rodwell and Hoskins (1996). Another effect is the enhanced upward motions over southeast of Japan. This area corresponds to the Baiu precipitation area. The downward motion in the western arid region

and the upward motion in the eastern wet region are associated with low-level heat flow created over the Eurasian continent in the mid-latitudes. Cold northerly flow in the western parts of the heat low tends to be balanced by the adiabatic warming due to downward motion. Warm southerly flow in the eastern region of the heat low is accompanied by the adiabatic cooling due to upward motion. These features are typically seen in the simulation with a concentrated deep heat source. Figures 5a-5c show the model response to a deep heat source cen-

Fig. 2. Observed streamfunction (106m2/s) for June at (a) upper-level and (b) low-level. (c) Observed vertical velocity (0.01Pa/s) at 500hPa for June. Shading is for less than -0.02Pa/s. Contours are for -0.08 -0.04, -0.02, 0.02, 0.04, and 0.08Pa/s. (d)-(f) are the same as (a)-(c) except for the simulated ones (EXP-JUN).

(a) PSI(200hPa) (*1.e6m2/s) JUN

(b) PSI(SIG=0.78) (*1.e6m2/s) JUN

(c) W(500hPa) (*0.01Pa/s) JUN

(d) PSI(200hPa) (*1.e6m2/s) EXP-JUN

(e) PSI(SIG=0.78) (*1.e6m2/s) EXP-JUN

(f) W(500hPa) (*0.01Pa/s) EXP-JUN


tered at (25N, 90E) under no-mountain condition (EXP-25N). Almost the same simulation is shown by Rodwell and Hoskins (1996). The heat source for EXP-25N has a Gaussian distribution with a longi-tudinal standard deviation of 10 and a latitudinal one of 5. The heating rates at the center of the heat source are 1.0, 3.0, 5.0, 3.0, and 1.0K/day for the upper five layers, respectively. It produces a heat low even in the mid-latitudes with a strong upward motion area on the northeastern side and a rela-tively weak but wide downward motion area on the northwestern side. That heat source creates an up-per anti-cyclonic circulation similar to the Tibetan High. The response to the concentrated deep heat source

depends on its latitudinal location. The mid-latitude response is quite small when a deep heat source is located in low latitudes of (10N, 90E), as in the pre-monsoon season (Rodwell and Hoskins, 1996). As compared with the case of a tropical heat source, the magnitude of the upper-level meridional flow over a subtropical heat source needs to be large

for the planetary vorticity advection to compensate the stretching term in the vorticity balance. Kato (1989) found that the low-level poleward

flow associated with the Baiu is geostrophically en-hanced by the South Asia heat low, which is pro-duced by the active convection due to the South Asia monsoon. Kodama (1993) supports the result, and concludes that the monsoon convections are im-portant heat sources to form heat lows as well as land-surface heating. The deep heat sources in the SOUTH region pro-

duce heat lows on the spot. This tropical heat low creates poleward flow in south-east Asia as concluded by Kato (1989) and Kodama (1993). Those deep heat sources also tend to directly pro-

duce the mid-latitude circulation preferred for the

Baiu formation. When EXP-SOUTH is compared

with EXP-TOPOG, we find that southwesterly flows

and upward motion are formed southeast of Japan.

Those are so weak that an anti-cyclonic circulation

expands into the continent as compared with the

corresponding one in EXP-JUN. If moisture is as-

Fig. 3. (a)-(c) and (d)-(f) are the same as Figs. 2a-2c except for EXP-SOUTH and EXP-TOPOG, respectively. Contours are for -0.08 -0.04, -0.02, 0.02, 0.04, and 0.08Pa/s. Dark and light shadings are for less than -0.02Pa/s and negative values (upward motions), respectively in (c) and (f).

(a) PSI(200hPa) (*1.e6m2/s) EXP-SOUTH

(b) PSI(SIG=0.78) (*1.e6m2/s) EXP-SOUTH

(c) W(500hPa) (*0.01Pa/s) EXP-SOUTH

(d) PSI(200hPa) (*1.e6m2/s) EXP-TOPOG

(e) PSI(SI6=0.78) (*1.e6m2/s) EXP-TOPOG

(f) W(500hPa) (*0.01Pa/s) EXP-TOPOG


sumed, an active precipitation process could occur

southeast of Japan. The mid-latitude heating due

to the Baiu precipitation itself is also important, as

shown in the next subsection.

The deep heat sources in the SOUTH region pro-

vides eastern Asia with a wet climate. The basically

similar feature is found in EXP-25N, as an upward-motion area northeast of the deep heat source. It is indicated from Fig. 8 of Rodwell and Hoskins (1996) that this is attributed to the interaction of tonal mean field with the Rossby-type response to the deep heat source; low-level southerly flow is created near the heat source to keep the vorticity balance. The southerly flow advects warm air northward in north-south temperature gradient field. The warm air moves upward north of the heat source to keep the heat balance while it moves northeastward under the mid-latitude westerlies.

3.3 Response to mid-latitude heat sources The latitude-height section of Q1 along 150E is

shown in Fig. 6. The major heating in the mid-latitudes is found around 35N, with its center at 500hPa. This heating is associated with precipita-tion in the Baiu fronts. The effect of the mid-latitude heating can be

clearly found in the comparison of zonal wind be-tween EXP-JUN and EXP-SOUTH. Figures 7a-7c show the latitude-height section of zonal wind along 150E for the observation, EXP-JUN, and EXP-

Fig. 4. Shaded area is specified for EXP-SOUTH (SOUTH area).

AREA for EXP-SOUTH

Fig. 5. (a)-(c) are the same as Figs. 3a-3c except for EXP-25N. Contours are for -0.08 -0.04, -0.02, -0.01, 0.01, 0.02, 0.04, and 0.08Pa/s.

(a) PSI(200hPa) (*1.e6m2/s) EXP-25N

(b) PSI(SIG=0.78) (*1.e6m2/s) EXP-25N

(c) W(500hPa) (*0.01Pa/s) EXP-25N

Fig. 6. The latitude-height section of diabatic heating rate (K/day) along 150E for June. Shading is for more than 0.5

K/day.

Q1(*K/day) 150E JUN


SOUTH, respectively. The observed zonal wind has its maximum over 32m/s at 200hPa around 38N, and expands to lower levels at 32N. The jet axis slopes northward as height increases. It is apparent that zonal wind in EXP-JUN is more similar to the observed one than EXP-SOUTH. Upper-level tonal wind speed in EXP-JUN is more than 32m/s around 38N. Low-level zonal wind locates at 30N, and cre-

ates a jet core at 700hPa. On the other hand, the upper-level wind in EXP-SOUTH is relatively weak and locates further north of the observed latitude. No vertical decline of the jet is simulated. A similar result was obtained by Kodama (1997). The low-level jet around 700hPa is one of the

most striking features of the Baiu front (Ninomiya and Akiyama, 1992; Yamazaki and Chen, 1993). A low-level jet separated from upper jet is observed in one case (Matsumoto et al., 1971). Yamazaki and Chen (1993) studied the effect of the Baiu precip-itation on wind enhancement near the Baiu front, based on the two-dimensional quasi-geostrophic ap-proximation. Their conclusions are that the low-level westerly wind is enhanced by the vertical gra-dient of diabatic heating due to the Baiu precip-itation. The strengthening of the low-level anti-cyclonic shear to the south of the front is also ob-tained. They pointed out that the anomaly fields induced from the diabatic heating have some sim-ilarity with the observed surrounding large-scale features. The comparison between EXP-SOUTH and EXP-JUN supports their indication. In EXP-JUN, anti-cyclonic low-level circulation near Japan in EXP-SOUTH shifts southeastward to the sub-tropics. The latitude-height section of vertical motions at

150E are presented with arrows in Fig. 7. The ob-served upward motions are distributed so as to com-pensate positive heating within 30-40N. Upward motions are simulated in EXP-JUN similarly to the observed ones in that area. Those vertical motions are forced by the prescribed positive heating. How-ever, weak upward motions are also simulated within 30-40N of EXP-SOUTH where no heating is pre-scribed in the mid-latitudes. Those are not found in the specified zonal mean field (Fig. 1d). The south-ern deep convections in EXP-SOUTH produce weak upward motions at the latitudes similar to the ver-tical motions for the Baiu precipitation.

It is noted that the upward motions within 10-20N and 50-60N are unrealistically large. These facts may be related to the weak anti-cyclonic cir-culation over the North Pacific in the simulations. Hoskins and Rodwell (1995) showed that longer in-tegrations are necessary to get the steady North Pa-cific high as compared with other circulations.

3.4 Response to near-surface heating Heat sources in the Asian summer monsoon is characterized by large and extended near-surface heating over the Eurasia Continent as well as the subtropical deep heat sources. We may have to

question how near-surface heat sources work on the monsoon circulation in the framework of the model response to the prescribed heat sources. Figure 8 shows the model response to near-surface

heat sources in June (EXP-SURF), where no dia-

Fig. 7. (a) Observed zonal wind (m/s) and vertical velocity (0.01Pa/s) along 150E for June are shown with contours and arrows, respectively. (b) and (c) are the same as (a) except for EXP-JUN and EXP-SOUTH, respectively. Shading is made for more than 4m/s. Arrows for less than 0.01Pa/s in magnitude are omitted.

(a) U(m/s) and W(*0.01Pa/s) 150E JUN

(b) U(m/s) and W(*0.01Pa/s) EXP-JUN

(c) U(m/s) and W(*0.01Pa/s) EXP-SOUTH


basic heating is given at any levels except the two lowest levels. The upper-level response is weak, and the Tibetan High and the mid-Pacific trough are not simulated. The low-level response is basically sim-ilar to the response in EXP-TOPOG, except that anti-cyclonic circulations are strengthened at the low-level from northern India through East Asia to the Pacific, as compared with EXP-TOPOG. We can see that the model response in 500hPa verti-cal motion is also strengthened particularly over the western arid region. Hoskins and Karoly (1981) discussed the circula-

tion response to various thermal heat sources. A shallow heat source in the mid-latitudes tends to be balanced by cold air advection due to northerly low-level flow. Planetary vorticity advection due to the northerly flow is compensated by the shrinking of the air column due to downward motion aloft. The differences of the results of EXP-SURF from those of EXP-TOPOG are roughly consistent with this theory. Northerly flows and downward motions are strengthened around the eastern Mediterranean Sea and in central Asia. Upward motions are inten-sified over the southerly wind area east of the heat

lows - the Caspian Sea and the eastern side of the Tibetan Plateau. The near-surface heating contributes to downward

motions over the western arid region, particularly east of the Caspian Sea and west of the Tibetan Plateau, more than the southern deep heating does. Results in the next section indicate that the sim-ulated effect of near-surface heating is too strong. In addition, in the real world, large surface sensi-ble heat flux in the western arid regions is closely related to little rainfall and little cloudiness within downward motions due to the southern deep heat-ing. Low-level circulations produced solely by near-

surface heating do not resemble the observed low-level monsoon circulation. Cross-equatorial flow, monsoon trough, and the low-level westerly jet in South Asia are not simulated at the appropriate lo-cation. This deficiency in EXP-SURF is more satis-factorily simulated by the southern deep heat source in EXP-SOUTH (Fig. 3b). The deep convective heating over the SOUTH region is the direct forcing for the South Asian summer monsoon circulation.

4. Middle Summer (July)

4.1 Heat sources and circulation Figures 9a and 9b are the mass-weighted mean Q1

between 85 hPa and surface for July and its differ-ence from that for June. The basic features of Q1 in June is maintained in July. We can see that deep heat sources over the Arabian Sea and the Bay of Bengal tend to move inland of India. The heating due to the Baiu precipitation shifts northwestward from south of Japan. Figures 9c and 9d are the same as Figs. 9a and

9b except for the climatological precipitation. The seasonal change of the climatological precipitation from June to July is well captured by that of the 1985-87 Q1 as well as the July distribution itself. The specified zonal mean field and its seasonal

change from June to July are shown in Figs. 9e and 9f, respectively. Since the northern troposphere be-comes warm entirely as compared with that in June, the subtropical jet moves northward and becomes weak. Figures 10a and 10b show the observed low-level

streamfunction and vertical velocity at 500hPa for July. Figures 10c and 10d are the same as Figs. lOa and 10b, respectively, except for the differences between July and June. Low-level anti-cyclonic changes are found over the Arabian Sea and over the East China Sea. Those anti-cyclonic changes mean the northward shift of low-level jet in South Asia and East Asia, respectively. Low-level negative streamfunction changes are found in the northern central Pacific. Upward motion change is extended from north of the Bay of Bengal through Japan. An-other one is distributed from (20N, 150E) north-

Fig. 8. (a)-(c) are the same as Figs. 3a-3c except for EXP-SURF.

(a) PSI(200hPa) (*1.e6m2/s) EXP-SURF

(b) PSI(SIG-0.78) (*1.e6m2/s) EXP-SURF

(c) W(500hPa) (*0.01Pa/s) EXP-SURF


eastward. The drastic change of vertical motion in East Asia corresponds to the northwestward shift of the Baiu precipitation zone. Roughly, upward mo-tion tends to move from the ocean side of the coasts to the continental side in East Asia and South Asia. Downward motion in the western arid region shifts its center northwestward and tends to be weak in the Middle East and northern Africa. The simulations for July (EXP-JUL) and their

differences from the June simulations (EXP-JUN) are shown in Figs. 11a-11b and 11c-11d, respec-tively. The details in the EXP-JUL simulation are not necessarily comparable to the July observation. As mentioned in Subsection 3.2, the deficiency in the North Pacific may be attributed to the integration periods. However, some characteristics in the sea-sonal change from June to July are simulated; low-level anti-cyclonic changes are simulated in South Asia and East Asia. Cyclonic changes are found in the subtropical Pacific. The large-scale pattern

of the observed vertical motion change is roughly captured by the corresponding differences between EXP-JUL and EXP-JUN.

4.2 Effect of zonal mean field change The differences between EXP-JUL and EXP-JUN

are attributed to the seasonal changes of Q1 and zonal mean field from June to July. Figure 12 shows the differences of the simulated circulation between EXP-JULWHOLE and EXP-JUN. The July tonal mean field and the June Q1 on the globe are used in EXP-JULWHOLE. The result (Fig. 12a) is interest-ing since the model response to the July zonal mean field seems to capture some qualitative characteris-tics of the monsoon circulation change from EXP-JUN to EXP-JUL (Fig. 11c); low-level anti-cyclonic changes are found along the coasts in East Asia and South Asia so that low-level jets tend to flow inside of the continent. The characteristic low-level jets in June both over South Asia and on the southeastern

Fig. 9. (a), (c) and (e) are the same as Figs. la, lc and id, respectively, except for July. (b), (d) and (f) are the same as (a), (c) and (e) except for the differences between July and June. Contours in (b) are for -3.0, -2.0, -1.0, -0.5, 0.5, 1.0, 2.0 and 3.0K/day. The dark and light shaded areas are for

more than 0.25K/day and less than -0.25K/day, respectively. Contours in (d) are for -8.0, -4.0, -2.0, -1.0, 1.0, 2.0, 4.0, and 8.0mm/day. Shading is for more than 1.0mm/day. Shading in (f) is for positive values.

(a) Q1 (*K/day) JUL (b) Q1 (*K/day) JUL-JUN

(c) Precipitation (mm/day) JUL (d) Precipitation (mm/day) JUL-JUN

(e) U(m/s) and W(*0.01 Pa/s) ZMEAN JUL (f) U(m/s) ZMEAN JUL-JUN


side of the Baiu precipitation are weakened. The differences of vertical motion between EXP-

JUL and EXP-JUN (Fig. 11d) are quite similar to those between EXP-JULWHOLE and EXP-JUN (Fig. 12b). We can confirm that the qualitatively similar pattern is also found in the observed sea-sonal change from June to July (Fig. 10d). The differences between EXP-JULWHOLE and

EXP-JUL (the differences between Fig. lid and Fig. 12b) are considered as the effect of the Q1 change. The changes of the vertical motion due to the zonal mean field (Fig. 12b) are consistent with the change of Q1 (Fig. 9b). The positive/negative Q1 change produces the changes toward upward/downward motions on the spot; upward motion tendency over India, the western Pacific, the Philippines and a

Fig. 10. (a) and (b) are the same as Figs. 2b and 2c except for July. (c) and (d) are the same as (a) and (b) except for the differences between July and June. Contours are -0.04, -0.02, 0.02 and 0.04Pa/s in (d). Dark and light shadings are for less than -0.02Pa/s and negative values (upward motions), respectively in (d).

(a) PSI(SIG=0.78) (*1.e6m2/s) JUL

(b) W(500hPa) (*0.01Pa/s) JUL

(c) PSI(SIG=0.78) (*1.e6m2/s) JUL-JUN

(d) W(500hPa) (*0.01Pa/s) JUL-JUN

Fig. 11. The same as Fig. 10 except for the simulated ones (EXP-JUL and EXP-JUN).

(a) PSI(SIG=0.78) (*1.e6m2/s) EXP-JUL

(b) W(500hPa) (*0.01Pa/s) EXP-JUL

(c) PSI(SIG=0.78) (EXP-JUL)-(EXP-JUN)

(d) W(500hPa) (EXP-JUL)-(EXP-JUN)


downward motion tendency over the Arabian Sea, the Bay of Bengal, and north of the Philippine Sea. It is interesting that most of these tendencies of the vertical motions do not contradict the changes due to the tonal mean field. This fact indicates that the convective activities distribute consistently with zonal mean fields in the climatological seasonal change. The effects of the seasonal change of the zonal

mean field are understood to some extent by con-sidering the interaction of tonal mean field change with the southern deep heating and the near-surface heating separately. Figure 13a shows the differences of vertical veloc-

ity between EXP-JULSOUTH and EXP-SOUTH, where EXP-JULSOUTH is the same as EXP-SOUTH except that the zonal mean field for June is replaced by one for July. The differences of vertical velocity between EXP-JULWHOLE and EXP-JUN (Fig. 12b) are captured well over south-east Asia by the differences between EXP-JULSOUTH and EXP-SOUTH (Fig. 13a) - upward motion change from India through the northern Arabian Sea and downward motion change in the southern Arabian Sea. The similarity is also found to some extent over East Asia and the Pacific. The effect of zonal mean field change is more

clearly shown in the model response to a local deep heat source. Figure 13b shows vertical velocity for EXP-JUL25N, where EXP-JUL25N is the same as EXP-25N except that the July zonal mean field is prescribed. The model response in the mid-latitudes

is weak in the magnitude as compared with EXP-25N (Fig. 5c). The horizontal extension of vertical motions is also different from that in EXP-25N. The upward motion branch expands northward from the heat source around (90E, 25N) in EXP-JUL25N rather than northeastward in EXP-25N. Downward motion on the western side of the heat source is shifted northwestward. The upward motion change over the northern Ara-

bian Sea and northern India from EXP-SOUTH to EXP-JULSOUTH (Fig. 13a) can be identified with the upward motion tendency north and northwest of the heat source in EXP-JUL25N (Fig. 13b) as com-pared with EXP-25N (Fig. 5c). The downward mo-tion change over the Baiu area from EXP-SOUTH to EXP-JULSOUTH (Fig. 13a) may be related to the differences of vertical motions in the northeast side of the heat source between EXP-JUL25N (Fig. 13b) and EXP-25N (Fig. 5c). To identify them, fur-ther specific experiments are necessary.

Fig. 12. (a) and (b) are the same as Figs. 10c and 10d except for the differences between EXP-JULWHOLE and EXP-JUN.

(a) PSI(SIG=0.78) (EXP-JULWHOLE)-(EXP-JUN)

(b) W(500hPa) (EXP-JULWHOLE)-(EXP-JUN)

Fig. 13. The same as Fig. lOd except for (a) the differences between

EXP-JULSOUTH and EXP-SOUTH and (c) the differences between EXP- JULSURF and EXP-SURF. (b) is the same as Fig. 5c except for EXP-JUL25N.

(a) W(500hPa) (EXP-JULSOUTH)-(EXP-SOUTH)

(b) W(500hPa) EXP-JUL25N

(c) W(500hPa) (EXP-JULSURF)-(EXP-SURF)


Next, the interaction of zonal mean field change with the near-surface heating is studied. Figure 13c shows the EXP-JULSURF, which is the same as EXP-SURF except that June zonal mean fields are replaced by the July ones. The differences of vertical velocity between EXP-JULSURF and EXP-SURF (Fig. 13c) are similar to those between EXP-JULWHOLE and EXP-JUN in the western arid re-gion (Fig. 12b). This suggests that the weakened dynamical effect of topography in July is more im-portant than the seasonal change of the near-surface heat source. However, the existence of near-surface heating is involved in that interaction. This is confirmed from the fact that the corresponding re-sults are not found in the differences between EXP-JULSOUTH and EXP-SOUTH (Fig. 13a) in the mid-latitudes. It is indicated from the observed vertical velocity change in the mid-latitudes (Fig. 10d) that the simulated response due to near-surface heating in EXP-JUN and EXP-JUL (Fig. 11d) is too strong. To summarize this study, the seasonal change

from June to July is characterized by the weakened interaction between the subtropical jet and the deep

and near-surface heating.

5. Late Summer (August)

5.1 Heat sources and circulation Figure 14 shows (a) the mass-weighted mean Ql

between 85 hPa and the surface for August and (b) its change from July. The basic features for the Asian summer monsoon are maintained in Q1 for August. The Q1 decreases over the northern conti-nents. A deep heating source over the Bay of Bengal shifts southward. The most drastic change of Q1 is found over the subtropical western North Pacific. Large deep heating is extended from the Philippines eastward to the dateline over the western Pacific. These characteristic seasonal changes of deep heat sources are confirmed by the climatological precipi-tation shown in Figs. 14c and 14d. According to Murakami and Matsumoto (1994) and Ueda and Yasunari (1996), the western warm SST and the associated SST gradient are basically responsible for the occurrence of deep convection over the western Pacific in late summer, at least in a climatological sense. The specified tonal mean zonal wind, vertical ve-

Fig. 14. The same as Fig. 9 except for August and July.

(a) Q1 (*K/day) AUG (b) Q1 (*K/day) AUG-JUL

(c) Precipitation (mm/day) AUG (d) Precipitation (mm/day) AUG-JUL

(e) u(m/S) and W(*0.01Pa/s) ZMEAN AUG (f) U(m/s) ZMEAN AUG-JUL


locity and their seasonal change from July to Au-gust are shown in Figs. 14e and 14f. The seasonal change of the tonal wind is small compared with that from June to July (Fig. 9f), especially in the mid-latitudes. Figures 15a-15b show the observed upper-level

streamfunction and low-level streamfunction for Au-gust. Figures 15c-15d are the same as Figs. 15a-15b except for the corresponding differences between August and July. Low-level heat lows over the Eurasian Continent are entirely weakened. The low-level westerly jet in South Asia shifts southward. The Tibetan High seems to shrink from the north-

western area of the Continent and expand to over Japan. Low-level westerly jet in South Asia tends to invade the western Pacific. Anti-cyclonic circula-tion prevails over Japan in upper- and low-levels. Figures 16a-16d are the same as Figs. 15a-15d

except for the simulated ones. The August simula-tion is referred to as EXP-AUG. Major features for the Asian summer monsoon in August are obtained. The seasonal change of circulation over Japan has a barotropic structure of anti-cyclonic circulations, as observed.

5.2 Effect of the western Pacific heat source Two regional parts of the August Q1 are consid-ered separately. The two regions are shown in Fig. 17. Those are referred to as the western Pacific (WP) region and the Indian Ocean and Eurasian Continent (EU) region, respectively. The EXP-WP case is the same as EXP-JUL except that the Q1 in July over the WP region is replaced by that of Au-

gust. The EXP-EU case is also the same as EXP-JUL except that the Q1 in July over the EU region is replaced by that of August. Zonal mean field is fixed to the July one in both cases. Figure 18 shows the differences between EXP-WP

and EXP-JUL in (a) the upper-level streamfunction and (b) the low-level streamfunction. The upper-level streamfunction in EXP-WP captures its major changes from EXP-JUL to EXP-AUG over the Pa-cific and even over the Indian Ocean. The organized heating over the WP region is responsible for that circulation response. Low-level circulation response tends to be inverse to the upper-level response in the tropical and subtropical regions. The response in the mid-latitudes, including Japan, is barotropic rather than baroclinic. The seasonal change of zonal mean component is not negligible for the low-level differences in high latitudes. Figures 18c and 18d are the same as Figs. 18a and

18b, respectively, except for EXP-EU. Cyclonic cir-culation anomalies over the Eurasian Continent and anti-cyclonic anomalies over Japan are simulated. The simulated low-level circulation in EXP-EU is limited around South Asia and is consistent with the southward shift of the South Asia monsoon flow.

5.3 Effect o f June zonal mean field The climatological seasonal change of zonal mean

field from July to August (Fig. 14f) is small com-pared with that from June to July (Fig. 9f). But the situation is different in the interannual van-abilities. For example, during the cool summer in Japan in 1993, zonal mean zonal wind keeps its loca-

Fig. 15. (a) and (b) are the same as Figs. 2a and 2b except for August. (c) and (d) are the same as (a) and (b) except for the differences between August and July. Dark and light shadings are for more than 2.106m2/s and less than -2.106m2/s in (c) and (d).

(a) PSI(200hPa) (*1.e6m2/s) AUG

(b) PSI(SIG=0.78) (*1.e6m2/s) AUG

(c) PSI(200hPa) (*1.e6m2/s) AUG-JUL

(d) PSI(SIG=0.78) (*1.e6m2/s) AUG-JUL


tion in lower latitudes than normal (e.g., Wakahara and Fujikawa, 1997) - positive anomalies around 200hPa and 35N and negative anomalies north of about 45N. This is similar to the seasonal change of the zonal mean zonal wind from June to July (Fig. 9f), though the magnitude of the 1993 sum-mer anomalies amounts to less than a third. Figures 19a and 19b show the differences of low-

level streamfunction and 500hPa vertical veloc-ity, respectively, between EXP-ZM and EXP-AUG, where EXP-ZM is the same as EXP-AUG except that tonal mean field in August is replaced by one for June. Cyclonic anomalies are found over East Asia. Easterly anomalies are found from Indochina through the South Asian coast to northern Africa. Upward motion anomalies, which imply enhanced precipitation, are simulated over Japan. Downward motion anomalies, which imply suppressed convec-

tive activities, are simulated near the Philippines and over South Asia. These results indicate that the zonal mean field delayed from its seasonal mi-gration could be related to weak monsoon circula-tion and the associated precipitation anomalies in the mid-latitudes and the subtropics.

6. Conclusions

The seasonal change of the Asian summer mon-soon circulation is studied, using the model with the prescribed zonal mean field and the prescribed diabatic heat source. The following conclusions are obtained by considering the Asian summer monsoon circulation as the model response.

(1) The model successfully simulates the major characteristics of the climatological seasonal change of the Asian summer monsoon circulation.

Fig. 16. The same as Fig. 15 except for EXP-AUG and EXP-JUL.

(a) PSI(200hPa) (*1.e6m2/s) EXP-AUG

(b) PSI(SIG=0.78) (*1.e6m2/s) EXP-AUG

(c) PSI(200hPa) (EXP-AUG)-(EXP-JUL)

(d) PSI(SIG=0.78) (EXP-AUG)-(EXP-JUL)

Fig. 17. Dark and light shaded areas are specified for EXP-WP (WP area) and EXP-EU (EU area), respectively.

AREA far EXP-WP & EU


(2) During the Asian summer monsoon season, the deep heat sources in south-east Asia reproduce the Tibetan High and the associated downward

motion in the western part of the Eurasian Con- tinent. Low-level circulation in South Asia can be simulated mostly by those deep heat sources.

(3) During the Asian summer monsoon season,

downward motions are also induced aloft by near-surface heating over central Asia. Near- surface heating can not solely reproduce the observed low-level circulation in South Asia.

(4) In early summer (June), the deep heat sources in south-east Asia tend to form southwesterly low-level flow and upward motion southeast of Japan. Those are considered to be the background for the Baiu in East Asia, as well as

heat lows created in south-east Asia.

(5) In early summer (June), the mid-latitude heat sources associated with the Baiu precipitation produce a low-level jet to the south and tend to locate low-level anti-cyclonic circulation further south.

(6) Seasonal change from early summer (June) to mid-summer (July) is characterized by air temperature increase in the whole Northern Hemi- sphere and northward shift of weakened westerly jet. The change of the zonal mean field from June to July in the model explains the

major characteristics of the climatological seasonal change from June to July; low-level jets and upward motion areas in South Asia and East Asia shift from the ocean side to the land side of the coasts.

(7) The vertical motion change due to the zonal mean field change is consistent with that due to the seasonal change of the heat sources from early summer (June) to mid-summer (July).

Fig. 18. (a) and (b) are the same as Figs. 15c and 15d except for the differences between EXP-WP and EXP-JUL. (c) and (d) are the same as (a) and (b) except for the differences between EXP-EU and EXP-JUL.

(a) PSI(200hPa) (EXP-WP)-(EXP-JUL)

(b) PSI(SIG=0.78) (EXP-WP)-(EXP-JUL)

(c) PSI(200hPa) (EXP-EU)-(EXP-JUL)

(d) PSI(SIG=0.78) (EXP-EU)-(EXP--JUL)

Fig. 19. (a) and (b) are the same as Figs. 10c and lOd except for the differences

between EXP-ZM and EXP-AUG.

(a) PSI(SIG=0.78) (EXP-ZM)-(EXP-AUC)

(b) W(500hPa) (0.01Pa/s) (EXP-ZM)-(EXP-AUG)


(8) Seasonal change from mid-summer (July) to late summer (August) is characterized by en-

hanced convective activity in the extended area of the subtropical western Pacific. The change of the heat source over the western Pacific solely explains the major characteristics of the climatological seasonal change from July to August, not only over the Pacific but also over the In- dian Ocean. The expansion of the Tibetan High at upper-level and the Pacific High at low-level over Japan is also simulated only by the seasonal change of the western Pacific heat sources.

(9) The seasonal change of the heat sources over the Eurasian Continent and South Asia from

mid-summer (July) to late summer (August) is related to the shrink of the Tibetan High from north and west of the Eurasian Continent and the southward shift of low-level westerlies in South Asia.

(10) The model simulation with the combination of the diabatic heat source for August and the zonal mean field for June could produce weak

monsoon circulation and the associated precipitation anomalies in the mid-latitudes and the subtropics, as compared with the climatological

August simulation.

We are interested in applying the method here to interannual variabilities of the Asian summer monsoon circulation. The important factors in the seasonal change are not necessarily important for its interannual variations. But the concept of ac-celerated and delayed progress of climatic seasonal cycle is used to understand anomalous climate in year-to-year variabilities. Park and Schubert (1997) show the similarities of the atmospheric circulation anomalies associated with the East Asian hot sum-mer in 1994 to the climatological July to August change.

A question to be raised is whether a set of ob-served zonal mean field and heat sources obtained from the observational data can reproduce the Asian summer monsoon circulation for each summer. It would also be interesting to see how heat source, the zonal mean field and the Asian summer mon-soon circulation are related each other.

Acknowledgments

We are grateful to Dr. Akio Kitoh, Dr. Masato Sugi and Dr. Mark Rodwell for their useful com-ments on the paper. We would like to thank anony-mous reviewers for their constructive comments. Thanks also go to our colleagues at the Meteoro-logical Research Institute (MRI) for their help and support for this study and the use of the MRI-GCM-I program code.

The figures are made with use of the GrADS (the Grid Analysis and Display System) provided by the COLA (Center for Ocean-Land-Atmosphere Stud-ies).

A part of this study is made in the Japanese Experiment on Asian Monsoon or JEXAM (1989-1998) supported by the Japan Science and Technol-ogy Agency.

References

Arakawa, A. and V. Lamb, 1977: Computational designof the basic dynamical processes of the UCLA gen-eral circulation model. Methods in ComputationalPhysics, Advances in Research and Application, Vol.17: General circulation models of the atmosphere,Academic Press, Inc., 337pp.

Ding, Y., 1992: Summer monsoon rainfalls in China.J. Meteor. Soc. Japan, 70, Special edition on Asiansummer monsoon, 373-396.

Hoskins, B.J. and D.J. Karoly, 1981: The steady linearresponse of a spherical atmosphere to thermal andorographic forcing. J. Atmos. Sci., 38, 1179-1196.

Hoskins, B.J., H.H. Hsu, IN. James, M. Masutani, PD.Sardeshmukh and G.H. White, 1989: Diagnostics ofthe global atmospheric circulation based on ECMWFanalysis 1979-1989. Department of Meteorology,University of Reading, Compiled as part of the U.K.Universities Global Atmospheric Modelling Project,WMO/TD, No. 326, 217pp.

Hoskins, B.J. and M.J. Rodwell, 1995: A model of theAsian summer monsoon. Part I: the global scale. J.Atmos. Sci., 52, 1329-1340.

Kato, K., 1989: Seasonal Transition of the lower-levelcirculation systems around the Baiu Front in Chinain 1979 and its relation to the northern summer mon-soon. J. Meteor. Soc. Japan, 67, 249-265.

Kodama, Y., 1992: Large-scale common features of sub-tropical precipitation zones (the Baiu frontal zone,the SPCZ, and the SACZ) Part I: Characteristics ofsubtropical frontal zones. J. Meteor. Soc. Japan, 70,813-836.

Kodama, Y., 1993: Large-scale common features of sub-tropical convergence zones (the Baiu frontal zone, theSPCZ, and the SACZ) Part II: Conditions of the cir-culations for generating the STCZs. J. Meteor. Soc.Japan, 71, 581-610.

Kodama, Y., 1997: The effect of the condensational heatsource in the subtropical precipitation zones on cir-culations (in Japanese). Abstracts for Japan meteo-rological Society Meeting, 72, 121.

Matsumoto, S., K. Ninomiya and S. Yoshizumi, 1971:Characteristic features of "Baiu" front associatedwith heavy rainfall. J. Meteor. Soc. Japan, 49, 267-281.

Murakami, T. and J. Matsumoto, 1994: Summer mon-soon over the Asian continent and western North Pa-cific. J. Meteor. Soc. Japan, 72, 719-745.

Ninomiya, K. and H. Muraki, 1986: Large-scale circula-tion over East Asia during Baiu period of 1979. J.Meteor. Soc. Japan, 64, 409-429.


Ninomiya, K. and T. Murakami, 1987: The early sum-mer rainy season (Baiu) over Japan. Monsoon Me-teorology, Oxford University Press, 93-121.

Ninomiya, K. and T. Akiyama, 1992: Multi-scale fea-tures of Baiu, the summer monsoon over Japan and the East Asia. J. Meteor. Soc. Japan, 70, Special edition on Asian summer monsoon, 467-495.

Park, C.K. and S.D. Schubert, 1997: On the nature of the 1994 East Asian summer drought. J. Climate, 10, 1056-1070.

Rodwell, M.J. and B.J. Hoskins, 1996: Monsoons and the dynamics of deserts. Quart. J. Roy. Meteorol. Soc., 122, 1385-1404.

Tanaka, M., 1992: Intraseasonal oscillation and the on-set and retreat dates of the summer monsoon over East, Southeast Asia and the western Pacific region using GMS high cloud amount data. J. Meteor. Soc. Japan, 70, Special edition on Asian summer mon-soon, 613-629.

Tokioka, T., 1978: Some considerations on vertical dif-ferencing. J. Meteor. Soc. Japan, 56, 98-111.

Tokioka, T., K. Yamazaki, I. Yagai and A. Kitoh, 1984: A description of the Meteorological Research In-stitute atmospheric general circulation model (The MRI-GCM-I). Tech. Rep. of the MRI. No. 13. 249pp.

Ueda, H., T. Yasunari and R. Kawamura, 1995: Abrupt seasonal change of large-scale convective activity over the western Pacific in the northern summer. J. Me-teor. Soc. Japan, 73, 795-809.

Ueda, H. and T. Yasunari, 1996: Maturing process of the summer monsoon over the western north Pacific -a coupled ocean/atmosphere system -. J. Meteor. Soc. Japan, 74, 493-508.

Wakahara, S. and N. Fujikawa, 1997: Study on abnor-mal climate in 1993/94 summer (in Japanese). Kisho Kenkyu Note. No. 189. 2-69.

Xie, P. and P. Arkin, 1996: Analysis of global monthly precipitation using gauge observations, satellite esti-mates, and numerical model predictions. J. Climate, 9, 840-858.

Yamazaki, N. and T.-C. Chen, 1993: Analysis of the East Asian summer monsoon during early summer of 1979: structure of the Baiu front and its relationship to large-scale fields. J. Meteor. Soc. Japan, 71, 339-355.

Yanai, M., S. Esbensen and J.-H. Chu, 1973: Determi-nation of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. At-mos. Sci., 30, 611-627.


夏のアジアモンス-ン循環と熱源の季節変化

尾瀬智昭

(気象研究所)

観測から得られた帯状平均場と熱源を与えたモデルを用いて、その応答としてのアジアモンスーンの循

環、特にその気候学的な季節変化を調べた。結果は次のようにまとめられる。

(1) 夏のアジアモンスーン期において、アジアの南域の深い積雲対流に伴う熱源は、チベット高気圧、モ

ンスーントラフならびに、南アジアにおける下層循環を形成し、さらにユーラシア大陸西部に下降流

をもたらす。中央アジアの地表面付近の熱源もまた、その上層に下降流を形成する。

(2) 初夏 (6月) に見られる、アジアの南域の深い対流に伴う熱源は、日本の南東に南西の下層風と上昇流

を形成する傾向を示す。これらは、アジアの南域に形成される熱的低圧部とともに、東アジアに梅雨

帯が形成される背景的要因になっていると考えられる。中緯度の梅雨帯での降水による熱源は、その

南に下層ジェットを形成する。

(3) 初夏 (6月) から盛夏 (7月) にかけてのアジアモンスーンの季節変化は、北半球全体で気温が上昇し、

帯状平均のジェットが北上し弱まることによって特徴づけられる。モデルにおいて帯状平均場のみを

6月から7月に変えると、この季節変化の主な特徴が再現される。すなわち、南アジアから東アジア

域において、下層ジェットおよび上昇流域は、大陸周辺の海洋から大陸側に移動する。鉛直流のこの

変化は、6月から7月にかけての熱源の季節変化と矛盾しない。

(4) 盛夏 (7月) から晩夏 (8月) にかけてのアジアモンスーンの季節変化は、亜熱帯西太平洋の広い範囲で

積雲活動が活発化することによって特徴づけられる。モデルにおいて西太平洋の熱源のみを7月から

8月に変えると、この太平洋域からインド洋域の季節変化の主な特徴が再現される。上層のチベット

高気圧および下層の太平洋高気圧が日本付近へ広がることもまた、西太平洋の熱源のみの季節変化で

再現される。

(5) 6月の帯状平均場と8月の熱源を用いたモデル実験が、気候学的な8月の実験結果と比べられる。気

候学的な季節変化から遅れた帯状平均場が、中緯度および亜熱帯域の弱いモンスーン循環および関連

する降水偏差と関連していることが暗示される。

Journal of the Meteorological Society of Japan, Vol. 76 ...

Documents

Transcript of Journal of the Meteorological Society of Japan, Vol. 76 ...